Skip to content

GLOSSARY.md

Latest commit

 

History

History
594 lines (395 loc) · 14.8 KB

GLOSSARY.md

File metadata and controls

594 lines (395 loc) · 14.8 KB

Glossary of Technical Terms

A comprehensive glossary of legal, technical, and domain-specific terms used in the Case Law Explorer project.

Legal Terms

ECLI (European Case Law Identifier)

A uniform identifier system for European case law. Format: ECLI:CountryCode:Court:Year:Ordinal

Example: ECLI:NL:HR:2023:1234

  • NL = Netherlands
  • HR = Hoge Raad (Supreme Court)
  • 2023 = Year
  • 1234 = Sequential number

Purpose: Enables unique identification and citation of court decisions across European jurisdictions.

Resources: Official ECLI website

CELEX (Common European Legal X-change)

A unique identifier for EU legal documents, including court cases. Format varies by document type.

Example for Court Case: 62019CJ0100

  • 6 = Case law
  • 2019 = Year
  • CJ = Court of Justice
  • 0100 = Case number

Purpose: Identifies all EU legal documents in the EUR-Lex database.

Resources: EUR-Lex CELEX info

BWB (Basis Wetten Bestand)

The official Dutch abbreviation for the "Basic Laws Database" - the official publication of Dutch laws and regulations.

Example: BWBR0001854 (Dutch Civil Code)

  • BWB = Basic Laws Database
  • R = Regulation
  • 0001854 = Sequential number

Purpose: Unique identification of Dutch legislation.

Resources: wetten.overheid.nl

Rechtspraak

Dutch term meaning "case law" or "jurisdiction". Refers to:

  1. The Dutch Judiciary (organization)
  2. The website rechtspraak.nl hosting Dutch court decisions
  3. Dutch case law in general

Purpose: Primary source for Dutch court decisions in this project.

Resources: rechtspraak.nl

LIDO (Linked Data Overheid)

"Linked Data Government" - A Dutch government initiative providing linked open data about legislation and case law.

Services:

  • Citation extraction between cases
  • References to legal provisions
  • BWB ID mapping
  • Legal provision aliases

Data Format: RDF/Turtle files with linked data

Resources: linkeddata.overheid.nl

ECHR (European Court of Human Rights)

International court established by the European Convention on Human Rights.

Jurisdiction: Human rights violations in Council of Europe member states

Database: HUDOC (Human Rights Documentation)

Case Identifier: Uses both ECLI and ItemID systems

Resources: echr.coe.int

CJEU (Court of Justice of the European Union)

The EU's highest court, interpreting EU law and ensuring uniform application.

Components:

  • Court of Justice (CJ)
  • General Court (GC)

Database: CELLAR (Central Repository for Legal Documents)

Case Identifier: Uses both ECLI and CELEX

Resources: curia.europa.eu

CELLAR

The Central Repository for Legal Documents of the EU Publications Office.

Contains:

  • EU legislation
  • Court of Justice case law
  • Parliamentary documents
  • Official Journal

Access: SPARQL endpoint for programmatic access

Resources: publications.europa.eu

Citation Network

A graph representation of how court cases reference each other.

Nodes: Individual court cases (identified by ECLI)

Edges: Citation relationships between cases

  • Outgoing: Cases cited by this case
  • Incoming: Cases that cite this case

Purpose: Enables network analysis, identifying influential cases, citation patterns, and legal doctrine development.

Legal Provision

A specific article, paragraph, or section of a law or regulation that is referenced in a court decision.

Example: "Article 1:1 BW" (Article 1:1 of the Dutch Civil Code)

Components:

  • Law identifier (e.g., BW, Sr)
  • Structural elements (Book, Title, Chapter)
  • Article number
  • Sometimes paragraph and sub-paragraph

Juriconnect (JCI)

A standard for unambiguous referencing of legal sources in the Netherlands.

Format: jci1.3:c:BWBR0001854&artikel=1

  • jci1.3 = Version
  • c = Type (c=current version)
  • BWBR0001854 = BWB ID
  • artikel=1 = Article 1

Purpose: Precise linking to specific parts of legal texts.

Technical Terms

ETL (Extract, Transform, Load)

A data pipeline pattern consisting of three stages:

  1. Extract: Retrieve data from source systems
  2. Transform: Clean, normalize, and structure data
  3. Load: Store data in target system (database, data warehouse)

In this project: Extract from rechtspraak.nl/HUDOC/CELLAR → Transform to unified format → Load to DynamoDB/S3

Apache Airflow

An open-source platform for orchestrating complex data workflows.

Key Concepts:

  • DAG (Directed Acyclic Graph): Workflow definition
  • Task: Individual unit of work
  • Operator: Template for a task type
  • Scheduler: Triggers DAGs based on schedule
  • Worker: Executes tasks

UI: Web interface at http://localhost:8080

Resources: airflow.apache.org

DAG (Directed Acyclic Graph)

In Airflow context: A collection of tasks with dependencies, representing a workflow.

Directed: Tasks have a defined order (A → B → C)

Acyclic: No circular dependencies (prevents infinite loops)

Example: Extract → Transform → Load

Task Group

In Airflow: A way to organize related tasks together for better visualization and management.

In this project: Monthly task groups for parallel processing of different time periods.

DynamoDB

AWS's fully managed NoSQL database service.

Data Model: Key-value and document database

Key Features:

  • Automatic scaling
  • High availability
  • Flexible schema
  • Fast performance

In this project: Stores case metadata with three tables (ECLI-based, CELEX-based, ItemID-based)

Resources: aws.amazon.com/dynamodb

S3 (Simple Storage Service)

AWS's object storage service.

Use Cases: Files, backups, data lakes, static websites

In this project: Stores full text of court decisions and graph data (nodes/edges)

Resources: aws.amazon.com/s3

GraphQL

A query language for APIs that allows clients to request exactly the data they need.

Advantages:

  • Request only needed fields
  • Single endpoint
  • Strongly typed
  • Self-documenting

In this project: Optional AppSync API for querying case law data

Resources: graphql.org

AppSync

AWS's managed GraphQL service.

Features:

  • Automatic schema generation
  • Real-time subscriptions
  • Built-in authentication
  • DynamoDB integration

In this project: Provides GraphQL API for querying case data

Resources: aws.amazon.com/appsync

Docker

Platform for developing, shipping, and running applications in containers.

Container: Lightweight, standalone, executable package including code, runtime, libraries, and dependencies.

In this project: Airflow and all dependencies run in Docker containers

Resources: docker.com

Docker Compose

Tool for defining and running multi-container Docker applications.

In this project: Orchestrates 6+ services (Airflow webserver, scheduler, worker, PostgreSQL, Redis, etc.)

Configuration: docker-compose.yaml file

Resources: docs.docker.com/compose

SPARQL

Query language for RDF (Resource Description Framework) data.

Purpose: Query linked data and semantic web resources

In this project: Used to query CELLAR database for CJEU cases

Example Query:

SELECT ?case ?date WHERE {
  ?case a cdm:judgment .
  ?case cdm:judgmentDate ?date .
}

Resources: w3.org/TR/sparql11-query

RDF (Resource Description Framework)

Standard model for data interchange on the web.

Structure: Triples (Subject-Predicate-Object)

Example:

<case123> <hasDate> "2023-01-01"
Subject   Predicate  Object

In this project: LIDO data is in RDF format

Turtle (TTL)

A textual syntax for RDF that is more compact and readable than XML.

Example:

@prefix ecli: <http://data.europa.eu/eli/ontology#> .

<http://example.org/case/1> 
    ecli:date "2023-01-01" ;
    ecli:court "Supreme Court" .

In this project: LIDO exports are in Turtle format (.ttl.gz)

Celery

Distributed task queue for Python.

Components:

  • Broker: Message queue (Redis in this project)
  • Worker: Executes tasks
  • Result Backend: Stores task results

In this project: Airflow uses Celery to distribute tasks across multiple workers

Resources: docs.celeryproject.org

PostgreSQL

Open-source relational database management system.

Features:

  • ACID compliant
  • Advanced SQL features
  • Extensible
  • Reliable

In this project:

  1. Airflow metadata database
  2. LIDO legal provisions storage

Resources: postgresql.org

Redis

In-memory data structure store used as database, cache, and message broker.

In this project: Celery message broker for Airflow task queue

Resources: redis.io

Pandas

Python library for data analysis and manipulation.

Key Features:

  • DataFrame data structure
  • Reading/writing various formats (CSV, JSON, SQL)
  • Data cleaning and transformation
  • Statistical analysis

In this project: Core library for data transformation phase

Resources: pandas.pydata.org

Boto3

AWS SDK for Python.

Purpose: Interact with AWS services programmatically

In this project: Used to:

  • Upload data to S3
  • Write to DynamoDB
  • Query AWS resources

Resources: boto3.amazonaws.com

Data Pipeline Terms

Extraction

First phase of ETL: Retrieving data from source systems.

In this project:

  • Rechtspraak: Download XML dumps, call metadata API
  • ECHR: Query HUDOC API
  • CJEU: Query CELLAR SPARQL endpoint
  • Citations: Query LIDO API

Output: Raw CSV/JSON files in data/raw/{date}/

Transformation

Second phase of ETL: Cleaning and normalizing data.

Operations in this project:

  1. Column name normalization
  2. Data type conversions
  3. XML/HTML parsing
  4. ECLI validation
  5. Citation structuring
  6. Duplicate removal

Output: Clean CSV files in data/processed/{date}/

Loading

Third phase of ETL: Storing data in target systems.

In this project:

  • Metadata → DynamoDB tables
  • Full text → S3 buckets
  • Graph data → S3 buckets
  • Legal provisions → PostgreSQL

Incremental Extraction

Extracting only new or changed data since the last extraction.

Method: Using date ranges (start_date, end_date parameters)

Benefits:

  • Faster processing
  • Lower resource usage
  • Reduced API calls

Parallel Processing

Executing multiple tasks simultaneously.

In this project: Monthly task groups process different months in parallel

Benefits:

  • Faster overall processing
  • Better resource utilization
  • Independent month processing

Idempotency

Property where repeated operations produce the same result.

Example: Re-running extraction for a month doesn't create duplicates

Implementation: Check if data exists before extraction

Project-Specific Terms

Month Directory

Organization pattern where extracted data is stored in directories by month.

Format: data/raw/YYYY-MM-DD/

Example: data/raw/2023-01-01/

Purpose:

  • Parallel processing
  • Easy reprocessing of specific periods
  • Clear data organization

Metadata Extraction

Process of extracting structured information about a case (not the full text).

Includes:

  • Case identifiers (ECLI, CELEX)
  • Court name
  • Decision date
  • Procedure type
  • Subject matter
  • Citations

Full Text Extraction

Process of extracting the complete text of court decisions.

Storage: JSON format in S3

Structure:

{
  "ecli": "ECLI:NL:HR:2023:1234",
  "full_text": "Court decision text...",
  "sections": {...}
}

Citation Extraction

Process of identifying and structuring references between cases and to legal provisions.

Sources:

  • LIDO API (primary)
  • Text parsing (secondary)

Output:

  • Incoming citations (cases citing this case)
  • Outgoing citations (cases cited by this case)
  • Legal provisions cited

Graph Data

Representation of cases and citations as network nodes and edges.

Nodes File (cellar_nodes.txt):

ECLI:NL:HR:2023:1234|Supreme Court|2023-01-01

Edges File (cellar_edges.txt):

ECLI:NL:HR:2023:1234|ECLI:NL:HR:2022:5678

Purpose: Network analysis and visualization

Acronyms Quick Reference

Acronym Full Name Category
AWS Amazon Web Services Technology
BW Burgerlijk Wetboek (Civil Code) Legal
BWB Basis Wetten Bestand Legal
CELEX Common European Legal X-change Legal
CJEU Court of Justice of the European Union Legal
DAG Directed Acyclic Graph Technology
DDB DynamoDB Technology
ECHR European Court of Human Rights Legal
ECLI European Case Law Identifier Legal
ETL Extract, Transform, Load Technology
EU European Union Legal
JCI Juriconnect Legal
LIDO Linked Data Overheid Legal
RDF Resource Description Framework Technology
S3 Simple Storage Service Technology
SPARQL SPARQL Protocol and RDF Query Language Technology
SQL Structured Query Language Technology
TTL Turtle (file format) Technology
UI User Interface Technology
URL Uniform Resource Locator Technology

Need More Information?

Last Updated: October 2025